78 THE ROYAL SOCIETY OF CANADA 



Lord Rayleigh* believed that discrepancies were still out- 

 standing in the action of the Radiometer and hoped that the memoirs 

 quoted above would be critically examined and the whole question 

 rediscussed. 



G. D. West^ has also examined the radiometer and using the 

 ideas of thermal transpiration developed by Reynolds, Sutherland 

 and Knudsen has shewn how to explain radiometer phenomena. 



Denoting the pivot friction by P (P is independent of speed), 

 the revolutions per minute by N, the gas friction by kN, the power 

 of the source (point) by / and the distance from the source to the 

 radiometer by D, we get, on equating the force responsible for the 

 motion to the forces opposing the motion 



— =P + kN 



If Di is the least distance at which we just get no motion 



whence • — = ^ + kN 



or — = -. +k'N. 



1 . 



Plotting ~ against N we should get a straight line. 



The author has carried out several experiments using as sources 

 iron balls heated red hot in the ffames of Meker burners. The fîames 

 by themselves have a temperature between 1600-1700°C., the balls 

 would of course be at least 500°C. lower than this. Two sizes of 

 balls were used of diameter 7.8 cm., and 4.2 cm. respectively. The 

 fîames practically enveloped the balls. 



Two types of Radiometer were employed. No. 1 has mica vanes 

 blackened on one side, No. 2 has aluminium vanes faced on one side 

 with mica. With No. 1 the blackened surfaces retreat from the 

 heated body, with No. 2 the polished metal retreats. 



Sometimes a Fery Radiation Pyrometer was set to view the ball 

 to test the constancy of its radiation, and sometimes also another 

 radiometer similar to No. 1 was set up at a fixed distance from the 

 ball to see whether the gradual heating of the glass envelope would 

 cause any variation in the speed. Apparently it doesn't. 



^Nature, Vol. 81, 1909, p. 69. 



sProc. Lond, Phys. Soc, Vol. XXXI, p. 278; Vol. XXXII, pp. 1G6, 222. 



